Blog: batteries

Disassembly of a Renault Traction Battery (Master, Kangoo, Zoe)

This will be an article on the complete disassembly of a Renault/LG Chem Traction Battery which will be used to power the Renault Twingo. Although this type of battery has been around since about 2013, there is not much information around on these units, although they are very well designed in my eyes.

This particular battery is from a 2020 Renault Master, the very same battery can also be found on the second series Renault Kangoo. It has a capacity of 36kWh, 33kWh usable. The battery is very similar to the one used in the widely popular first series Renault Zoe, just a few minor differences to be noted.

For all that follows now: Electricity kills! Please be sure to wear full PPE if you are doing similar things in your garage. I am wearing cotton long sleeve clothes, safety/ESD shoes, 1000V electrical safety gloves, helmet and visor in case of electric arc. Here, I am measuring the total voltage of the battery pack.

After removing the main cover of the battery, we first see the high voltage connections. The main connection to the car on the right side, the service disconnect on the left side.

These here (the two black lines with a connector in the middle of the picture) are the pilot lines, one for the service disconnect and one for the main connector. The pilot lines are monitoring if the connectors are closed and disarm the connectors (by opening the main contactor) in case the connectors are getting opened on a live battery.

Within the battery box, sitting right on top of the battery modules, we find a module with the precharge circuit:

Let's have a look inside this component!

If we open the protection, it gets more visible: The main 275A fuse on the battery side on top, the main contactor (grey) just below, the black precharge relay further down and the precharge resistor (marked with a number 9) and the current sensor below the orange cable connection on the bottom left side.

Here we can see the BMS computer, which includes the voltage sensing boards and the BMS system in one enclosure, similar to e.g. the Orion BMS. The battery itself consists of 12 modules with 8 cells (technically 16 cells but in a 2p8s configuration), so we have 96 wires only for the voltage sensing cables, plus additional temperature sensors and other signalling cables.

The wiring loom leads from the battery modules to the BMS computer.

The individual modules are connected in series by busbars, which are enclosed in an orange plastic cover.

Next, we are going to work on the live battery in order to disconnect the modules. Please be aware that this is 400V DC, so we need to make sure that we use insulated tools in addition to the PPE.

By removing the busbars (second picture) we are breaking the pack and therefore reducing the voltage to a safe(r) level.

Here we can see the disconnected modules in the pack. The metal surfaces are the minus and plus connections of the modules, further right the white connector holds two thermistors for temperature control and the 8 voltage sensing connections for the individual cells to the BMS computer.

Here's another view of the liberated and disconnected modules, still within the battery case:

And here we have the first modules removed from the battery box:

If we now go down to the single module, we can see the 8 individual cells, stacked on top of each other like chocolate bars (yes, I am Swiss ;-) !)

Each of these cells is made of 2 LG Chem (what I believe to be NCM 712) pouch cells, connected in parallel within the cell module. So in practice we have 16 pouch cells per module, the BMS of course only sees 8 cells.

So that's the disassembly basically. I will now continue with some voltage measurements and the connectors pin arrangements in order to document these things here.

This battery will be going into the Renault Twingo, so stay tuned how this will work out!

The ZEBRA Ni-NaCl Battery

Now that we have removed all the covers of the battery box, let's have a look at the molten-salt (Ni-NaCl) battery that was used on this Renault Twingo. It is a pity really that I have to remove it: These batteries would be very environmental friendly, as the components of the battery are widely available without major impacts on the eco-system such as extensive mining etc.

If you want to read more about these batteries, check this Wikipedia article.

Here you can see the main battery with the HV and communication connections on the right side.

There were a few tests for the use of these batteries in the automotive industry. The first series of the Ford/Pivco Think were also powered by these batteries. In order to work, these batteries need to be kept at an operating temperatures of about 300 degrees Celsius / 570 degrees Fahrenheit, in order to keep the salt crystals in a liquid form. The heat is generated primarily by the chemical reaction within the battery while charging/discharging. If the battery is left idle, an external power source needs to be applied in order to keep the chemical reaction alive.

In a car, this means that the car must be plugged in when leaving it idle for more than 24 hours, otherwise the salt would crystallise and become solid, which would mean that the car needs to be heated up with an external power source for about 30 hours in order to liquidify the salt before the battery is ready to operate again, and the car wouldn't be usable during that time.

Plugging in the car at home wouldn't generally be a huge problem, as it can be plugged into a normal household socket and it draws only a very small amount of energy, just to keep the chemical reaction alive. But of course the handling is a bit less carefree than with a traditional car, so that's maybe one reason why the Ni-NaCl battery never got a more widespread use in the automotive industry, which is a pity really. The second reason might have been that there are only two serious larger-scale manufacturers of these units globally, which resulted in a very limited support network for these batteries. Ask some early users of the Think vehicles for their experience, and they will all tell you the same: The battery was absolutely great as long as there were no problems. The smallest problem however more often than not meant the end of life of the battery (in the better case) or (in most cases) even the whole car, as nobody was able to actually diagnose or even fix these batteries. That's probably also the reason why most of the Think vehicles were gone pretty quickly, even in Norway where they had quite some success in the very beginning. The ones still on the road underwent a similar surgery as I am planning on my Renault Twingo actually.

And here's a better look at the connections:

And this is very interesting now: The technical specs of this battery block. As you can see, the capacity is approx. 20 kWh, which considering the battery dimensions is a surprisingly good density. And another thing is quite surprising: This block has been built quite recently if you look at the sticker, the manufacture date reads March 2016!

The problem with the battery is that there are plenty of errors on the system, indicating either a isolation problem or indeed a faulty cell. It is quite hard to get these things diagnosed as there is little knowledge around on these special batteries. The main use of this kind of batteries nowadays is grid storage, which makes a lot of sense especially also as a network buffer for renewables (e.g. solar), where you could collect and store energy from the sun during the day and then power your appliances at your house in the evening or at night.

The plan now is to see if Fiamm/Sonick are willing to diagnose the battery, and if there's a chance to bring it back to life I might indeed be using it in a future solar application!

CALB 6s2P Battery Modules

The CALB 6s2p modules are another good battery option for tight spaces and higher voltages.

These modules were actually my number one choice for the Renault Twingo as I would be able to perfectly fit 12 modules of these with cooling plates into the original battery box in two layers separated by a cooling plate. Here's a quick drawing I made just to see how they would fit:

Let's go into some technical details of these:

One module is 22.2V nominal and 100Ah, which translates into 2.2 kWh per module. To do the same calcualtions as with the other batteries, here's what we would get for the Twingo:

Voltage: 12 x 22.2V=266.4V which would be absolutely perfect as we have a target voltage of 270V.

Engergy: 12 x 2.2kWh = 26.4kWh total capacity (very good!), density 26.4kWh/116 Litres = 0.227 kWh/Litre of space in the battery box. This is slightly higher than the Samsung SDI modules which come in at 0.214 kWh/Litre but still lower than the Tesla modules that come in at 0.274 kWh/Litre.

There is a very good source in Europe for these modules: Zero EV in the UK are doing a brilliant job in configuring whole battery packs with cooling plates, piping, busbars and everything. Check out their Mazda MX-5 Miata Conversion with these very same battery modules on Youtube where you see how these packs are assembled:

They use 10 modules in the Video, two less than I would have put in my Twingo.

So why didn't I go ahead with these modules in the end? Reading the technical sheets of these batteries I noticed that these modules are meant to be used in a normal position or flipped to their side by 90 degrees, but NOT mounted upside down (flipped by 180 degrees)! Not exactly sure where the problem would be, but I am not a huge fan of shortcuts when it comes to battery safety. In my setup, I would have mounted 4 battery modules per cooling plate, and put the cooling plates into the box in a horizontal way. So the lower 2 modules would of course be upside down.

Zero EV cooling plates for CALB modules, taken from the Zero EV Webiste.

If I wouldn't have identified a different, equally good solution in terms of size and voltage, I would probably have gone ahead with these CALB modules, as they are fitting so perfectly both in terms of size and voltage!

Samsung SDI LX86 Ford PHEV modules

So after we discovered in the last post that the total Voltage of 6 Tesla modules would only be half of the required value for the Renault Twingo, let's have a quick look at a different option just for the fun of it:

I have been offered a set of brand new Samsung SDI LX86 modules. These are Li-Ion modules configured in 12s1p with a nominal voltage of 43.2V (so double the value of the Tesla modules, as these here are in 12s whereas the Tesla modules are 6s). The Total capacity of a module is 2.07 kWh, 12 kg each. These modules are used in Ford hybrid vehicles (PHEV) such as the Ford Kuga PHEV.

Picture courtesy of Eco Lithium BV

The module measurements are 355 x 151 x 108 mm. So again if we have a quick look at the battery box measurements (800 x 520 x 280 mm) we can see that it would be possible to fit 12 modules into the box.

If we make the same calculation like for the Tesla modules, we would put 12 of these modules in series, which would result in 12 x 43.2V=518.4V. As you can easily see, this is much too high, as we were looking for about 270V. So if with the Tesla modules we were only half the value, here we have nearly double the value! So we will need to find something in between.

Let's do a quick calculation on density and capacity: 12 x 2.07kWh=24.84 kWh total capacity, 24.84/116 Litres = 0.214 kWh/Litre of energy density in the battery box. The Tesla solution would have had a density of 0.274 kWh/Litre in this box, so these Samsung SDI modules are nearly 20% less dense than the Tesla modules in this configuration.

But of course our main problem is the voltage here. If it would be possible to create two strings of 6 modules (2p6s), we would be pretty perfect (6 x 43.2=259.2V). Unfortunately, splitting packs into parallel strings creates a whole lot of complexity on the BMS and safety side. So unfortunately I had to abandon this route, even though I would have got these modules at a very convenient price.

Maybe you are interested how cooling of these modules would work. As you might have learned on the previous post, the Tesla modules have an integrated cooling circuit. These modules here don't have any plumbing for cooling included, so here we would work with cooling plates. The modules would be "glued" to the cooling plate with a heat transfer film. Other modules such as e.g. CALB-modules work in a very similar way.

Picture courtesy of Eco Lithium BV

So again, today we learned a lot about a specific battery technology, without finding a solution for our problem.

Electric Renault Twingo: Tesla Modules ¦ EV Conversion

The TESLA model S battery is always the first choice in conversions for many reasons: Excellent energy density, good availability, integrated liquid cooling system and easy BMS options (recommended BMS system: SimpBMS, see documentation here. Here's how simple this BMS looks like:

The main disadvantage of the Tesla module is its packaging. It needs a watertight outer battery box, and needs to be mounted on rails.

One Module measures 690 x 315 x 80 mm. So if you have a quick look at the battery box measurements (800 x 520 x 280 mm), you will see that if we would pack the modules in a horizontal way, we would only have space for 3 modules (280/80), but if we would be able to mount them in a vertical way on the long side, we could theoretically fit 6 of them (520/80). But as you can easily see, this is not possible in practice, as we would have 35mm missing in height.

Although there would probably be a solution to that by modifying the battery box itself in height, we would face another problem: The nominal voltage of a Tesla moudle is 22.8 V. Putting 6 of them in series means that we can add the voltages of the 6 modules, so 6x22.8V=136.8V. So we only are at half of the target voltage of 270V, which of course isn't any good. Some people went ahead to modify the Tesla modules with some surgery on the internal configuration of the pack (making it 12s instead of 6s) to double the voltage in order to meet the target voltage (would be perfect in my example), but this isn't something for the faint-hearted, and there are some other implications such as different currents and some other risks. I wouldn't recommend doing it, and I only know very few projects in automotive settings that effectively went ahead with 12s Tesla modules.

One Tesla module has a capacity of 5.3 kWh, so 6 modules would be 31.8 kWh. If we calculate the energy density in our specific battery box in the Renault Twingo, we get 31.8/116=0.274 kWh/Litre. So let's keep this in mind when comparing it with other battery systems!

So unfortunately we have to rule out the normal Tesla module at this point - let's look at other options!